AI Lead Research
How to find better prospects before you write a single message — build an ICP, spot buying triggers, and score lead quality.
1. Start here
Most people use AI for lead research the wrong way: they ask for "a list of prospects," copy whatever comes back, and wonder why the outreach feels generic. The problem isn't that AI is bad at research — it's that the thinking step got skipped. If your ideal customer is vague, your AI research will be vague.
This guide shows you how to research better prospects before writing a single message — for founders finding early customers, owners hunting better-fit leads, and consultants done chasing random prospects.
2. The five layers of a good lead
A good lead has five layers: ICP fit (is this the type of person or business you actually serve?), a buying problem they realistically care about, a trigger (a reason now), the buying committee (who's actually involved in the decision), and research proof — evidence supporting your reason for reaching out.
The guide walks each layer with the difference between weak and strong versions — a weak ICP is a category; a strong ICP has business type, buyer role, pain pattern, urgency sign, and exclusions.
3. Reusable tools, not one-off lists
By the end you have three reusable tools: an ICP builder prompt, an account research workflow, and a lead-quality scorecard. And just as important: a clear line on what not to automate, so your outreach stays human where it counts.